Chris Coleman

I'm currently a second-year computer science PhD student at Northwestern University, doing research related to machine learning and natural language processing. I'm co-advised by Doug Downey and Kris Hammond. Broadly, I'm interested in pre-trained language models:

• Mechanisms and limitations in acquiring knowledge and adapting to novel domains
• Disentangling the role of pre-training objectives, corpora, and architectures
• Understanding (extended) pre-training with specific tasks (e.g. domain-adaptive pre-training, extreme multi-tasking, chain-of-thought data)

Since June 2022, I've also been working as a research intern for Semantic Scholar at the Allen Institute for Artificial Intelligence to recycle computation from massive pre-trained models within lightweight pre-trained models. Here, I focus on scientific text and tasks (mostly biomedical texts).

Beyond my primary research threads, I try to keep up with synthetic biology research. As an undergraduate, I worked with the Leonard Lab at Northwestern University, developing software to aid the design of mammalian genetic programs.

Some things that I enjoy outside of research: oil painting, generative art, graphic design, video game speedrunning, solving sudoku variants, hiking, playing soccer and watching Bundesliga.

Site last updated on Dec 2, 2022.